Speech encompasses more than mere words; it incorporates paralinguistic elements that reflect language, personality, intention, and emotion. The rapid advancements in machine learning and deep learning technologies have led to significant strides in Speech Emotion Recognition (SER) systems, designed to detect and classify emotions in human speech. However, SER systems remain a challenging research area, necessitating substantial amounts of ground truth data for effective training and achieving satisfactory performance. To address these challenges, this study introduces a gamified application, augmented by crowdsourcing techniques, aimed at enhancing and apply SER models through the collection of annotated data. Specifically, the "J-PLUS" app offers three distinct modules tailored for learning and practicing emotional speech, catering to professional journalists, news anchors, and the general public. These modules encompass a serious game for interactive practice, authentic news content for professional education, and an emotional journal for personal expression. The overarching objectives of the application include enhancing user engagement, optimizing user experience, and supporting continuous practice and training endeavors.
(See document for exact affiliation information.)
https://aes2.org/publications/elibrary-page/?id=22590
(718KB)
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member Join the AES. If you need to check your member status, login to the Member Portal.
